File-based configuration

Hawk server includes an API to add Hawk instances that are used to index and query models. The configuration engine allows the server to create and configure Hawk instances as per user-created configuration files. The server should be ready to receive user queries upon startup without any interaction from user or clients. Upon startup, Hawk server reads and parses configuration files, and then it creates/updates hawk instances as per configuration files.

NOTE: the Hawk server no longer writes to configuration files. If an instance configuration changes during operation, this configuration is persisted through the current HawkConfig mechanism. Configuration files will not overwrite any of the changed settings. The only exception is the polling min/max which will revert to config file settings if a server is restarted.

Format

Configuration files are XML files that define hawk instance name and its configuration. An XML schema can be found at HawkServerConfigurationSchema.xsd. A sample configuration file can be found at Sample Configuration File

The XML should include the following elements:

Table 1: List of XML elements in configuration file

Element Name	Parent Element Name	multiplicity	Value	Description
‘hawk’	xml	1	None	Root element
‘delay’	‘hawk’	1	None	Polling configuration
‘plugins’	‘hawk’	0-1	None	List of plugins (to be/that are) enabled
‘plugin’	‘plugins’	0-*	None	Plugin name
‘metamodels’	‘hawk’	0-1	None	List of metamodels (to be/that are) registered
‘metamodel’	‘metamodels’	0-*	None	Metamodel parameters
‘repositories’	‘hawk’	0-1	None	List of repositories (to be/that are) added
‘repository’	‘repositories’	0-*	None	Repository parameters
‘derivedAttributes’	‘hawk’	0-1	None	List of derived attributes (to be/that are) added
‘derivedAttribute’	‘derivedAttributes’	0-*	None	Derived attribute parameters
‘derivation’	‘derivedAttribute’	0-1	None	Derivation parameters
‘logic’	‘derivation’	0-1	CDATA section	An executable expression of the derivation logic in the language specified.
‘indexedAttributes’	‘hawk’	0-1	None	List of indexed attributes (to be/that are) added
‘indexedAttribute’	‘indexedAttributes’	0-*	None	Indexed attribute parameters

Table 2: ‘hawk’ attributes

Element Name	Attribute name	Optional/Required	Type	Description
‘hawk’	‘name’	Required	String	The unique name of the new Hawk instance
	‘backend’	Required	String	The name of the backend to be used (e.g.org.eclipse.hawk.orientdb.OrientDatabase, org.eclipse.hawk.orientdb.RemoteOrientDatabase)

Table 3: ‘delay’ attributes

Element Name	Attribute name	Optional/Required	Type	Description
‘delay’	‘min’	Required	String	Minimum delay between periodic synchronization in milliseconds
	‘max’	Required	String	Maximum delay between periodic synchronization in milliseconds (0 to disable periodic synchronization)

Table 4: ‘plugin’ attributes

Element Name	Attribute name	Optional/Required	Type	Description
‘plugin’	‘name’	Required	String	e.g. (org.eclipse.hawk.modelio.exml.listeners.ModelioGraphChangeListener, org.eclipse.hawk.modelio.exml.metamodel.ModelioMetaModelResourceFactory, org.eclipse.hawk.modelio.exml.model.ModelioModelResourceFactory)

Table 5: ‘metamodel’ attributes

Element Name	Attribute name	Optional/Required	Type	Description
‘metamodel’	‘location’	Optional	String	Location of metamodel file to be registered
	~~‘uri’~~	~~Optional~~	~~String~~	~~Metamodel URI. This value is set automatically by server to list registered metamodels~~

Table 6: ‘repository’ attributes

Element Name	Attribute name	Optional/Required	Type	Description
‘repository’	‘location’	Required	String	Location of the repository
	‘type’	Optional	String	The type of repository available repository types ()
	‘user’	Optional	String	Username for logging into the VCS
	‘pass’	Optional	String	Password for logging into the VCS
	‘frozen’	Optional	String	If the repository is frozen (true/false)

Table 7: ‘derivedAttribute’ attributes

Element Name	Attribute name	Optional/Required	Type	Description
‘derivedAttribute’	‘attributeName’	Required	String	The name of the derived attribute
	‘typeName’	Required	String	The name of the type to which the derived attribute belongs
	‘metamodelUri’	Required	String	The URI of the metamodel to which the de- rived attribute belongs
	‘attributeType’	Optional	String	The (primitive) type of the derived attribute
	‘isOrdered’	Optional	String	A flag specifying whether the order of the values of the derived attribute is significant (only makes sense when isMany=true)
	‘isUnique’	Optional	String	A flag specifying whether the the values of the derived attribute are unique (only makes sense when isMany=true)
	‘isMany’	Optional	String	The multiplicity of the derived attribute

Table 8: ‘derivation’ attributes

Element Name	Attribute name	Optional/Required	Type	Description
‘derivation’	‘language’	Required	String	The language used to express the derivation logic. Available labguages in Hawk: org.eclipse.hawk.epsilon.emc.EOLQueryEngine, org.eclipse.hawk.orientdb.query.OrientSQLQueryEngine, org.eclipse.hawk.epsilon.emc.EPLQueryEngine

Table 9: ‘indexedAttribute’ attributes

Element Name	Attribute name	Optional/Required	Type	Description
‘indexedAttribute’	‘attributeName’	Required	String	The name of the indexed attribute.
	‘typeName’	Required	String	The name of the type to which the indexed attribute
	‘metamodelUri’	Required	String	The URI of the metamodel to which the indexed attribute belongs.

Location

Configuration files are expected to be located in the ‘configuration’ folder in the server’s home directory. Each Hawk instance should have its own configuration file. There are no rules on how the file should be named. It is a good practice to include hawk instance name in the file name for easy recognition.

How to use/enable Hawk instance configuration engine

You can follow this video tutorial, or alternatively follow these steps:

1. Download the hawk-server-*.zip file for your operating system and architecture of choice from Hawk Server With Configuration)
2. Create a configuration file for each instance required to run in the Hawk server.
3. Edit configuration files:
4. Set instance name, backend, delay
5. Add list of plugins to be enabled
6. Add metamodel file to location to be registered
7. Add repositories that are to be indexed
8. Add any required derived attributes
9. Add any required indexed attributes
10. Save the configuration files to the ‘configuration’ folder in the server’s home directory (see figure 1)
11. Perform any other configuration that are required by Hawk Server and start the server, by following the deployment instructions].
12. Check if the hawk instances are added and running by typing ‘hawkListInstances’ in the server’s command terminal:

Usage Notes

• Deleting configuration files from the directory will not delete instances from the server. However, the server will not start those instances.

• To test Hawk server with Measure Platform, refer to Using HawkQueryMeasure to query Hawk instance running in Hawk Server